When converting a dataframe to a dictionary, how do I extract only part of the dataframe to be the key?

Question

Here is what the dataframe looks like. It was retrieved from Santiment.

                           ethSpent
datetime                           
2020-04-17 00:00:00+00:00       0.0
2020-04-17 01:00:00+00:00       0.0

df.todict()

Here is what one pair of key and value looks like

Timestamp('2021-04-14 20:00:00+0000', tz='UTC'): 0.0,

How do I only extract the datetime without the Timestamp and tz='UTC'?

Answer 1

You could do it like this without changing the original df.

df = pd.DataFrame({
    'date' : pd.date_range('1/1/2022', '1/5/2022', freq='1D', tz='UTC'),
    'value' : list(range(5))
})
df.assign(data=lambda x: x['date'].dt.date)[['data','value']].to_dict()

Output:
{'data': {0: datetime.date(2022, 1, 1),
  1: datetime.date(2022, 1, 2),
  2: datetime.date(2022, 1, 3),
  3: datetime.date(2022, 1, 4),
  4: datetime.date(2022, 1, 5)},
 'value': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4}}

Like in the comments asked. You can just go with .reset_index() and then assigning like before:

df.reset_index().assign(data=lambda x: x['date'].dt.date)[['data','value']].to_dict()